The call for reforming police in the USA has brought support from Amazon, historically when the facial recognition software was released for police use. It has been criticised extensively because “It regularly falsely [identifies] Black and Brown people as criminal” according to Congresswoman Alexandria Ocasio-Cortez who believes it shouldn’t be anywhere near law enforcement. I was also shocked to learn that an experiment in 2018 run by the ACLU proved just how inaccurate the software was by incorrectly matching 28 members of Congress with photos of people who were arrested for committing a crime and was overwhelmingly inaccurate when it came to matching whose members who are not white. Facial recognition software, like many forms of artificial intelligence, has a long history of racial bias. The field of artificial intelligence, which is overwhelmingly white and male, is frequently criticised for its lack of diversity.
Amazon have now banned the use of it by law enforcement by implementing a one year moratorium. There are calls for Amazon to use the year off to improve the software and lobby congress for industry friendly regulation to ensure that people are not being wrongly identified or accused. I suspect the best place to start would be to make it accurate as given how incorrect it is i’m surprised it was allowed to be used after the ACLU data was made public. The injustice done since its implementation is immeasurable, it is horrible to think about the number of people incorrectly charged for crimes they haven’t even committed and yet the perpetrator continues to run free.
If you would like to grab a copy of my book and support my writing it is available in both print and kindle version check out the links below. Donate to your favourite charity when you buy my book using Amazon Smile!